Temporal and Spatial Conventions¶

Seasons¶

Data is organized into quarterly seasons formatted as YYYY_QN:

Season	Notes
`2019_Q4`	Pre-COVID baseline
`2021_Q2`, `2021_Q4`	Post-COVID recovery
`2022_Q4`
`2023_Q2`, `2023_Q4`
`2024_Q2`, `2024_Q4`
`2025_Q2`, `2025_Q4`	Current/upcoming

Each season has a validity start date that determines the specific week used for transit routing. The Thursday and Saturday of this week represent weekday and weekend travel respectively.

Days of Week¶

Places pipeline runs are built per day: - Thursday — represents typical weekday travel - Saturday — represents typical weekend travel

Regions¶

Hierarchy¶

Level	Examples	Use
Nationwide	`usa`	Full US builds, nationwide sources
Megaregion	`california_nevada`, `mid_atlantic`, `great_lakes`, `south_atlantic`, `southeast`, `texas`, `mountain_west`, `northwest`, `midwest`, `new_england`, `new_york_new_jersey`	Regional Places pipeline runs
Test	`mini_kc`, `mini_nor_cal`	CI testing, development iteration

Region ID Format¶

replica:region:{region_name}:{hash}:region

Look up region IDs in model-159019.geos.region.

Spatial Filtering Pattern¶

All regional data cutouts use the same pattern:

SELECT data.*
FROM `model-159019.{dataset}.{table}` AS data
JOIN `model-159019.geos.region` AS region
  ON region.id = '{region_id}'
  AND ST_COVERS(region.geom, data.geom)

For lat/lng columns instead of geometry:

JOIN `model-159019.geos.region` AS region
  ON ST_COVERS(region.geom, ST_GEOGPOINT(data.lng, data.lat))
WHERE region.id = '{region_id}'

H3 Hexagons¶

H3 hexagonal grid cells are used for gridded spatial analysis (geohash/hexbin). The routable_cells table in the transportation network CDP identifies H3 cells accessible by transit.

FIPS Codes¶

Census geography identifiers used throughout population and land use data: - State FIPS: 2-digit (e.g., 06 = California) - County FIPS: 5-digit state+county (e.g., 06037 = Los Angeles County) - Census Tract: 11-digit - Census Block: 15-digit